#import and load the package 
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from sklearn.ensemble import IsolationForest
import statsmodels.api as sm
import seaborn as sns
from sklearn.experimental import enable_iterative_imputer
from sklearn.impute import IterativeImputer
from scipy import stats
from scipy import stats
from sklearn.preprocessing import MinMaxScaler
from sklearn.decomposition import PCA
from sklearn.decomposition import NMF
from scipy.cluster.hierarchy import linkage, cophenet
from scipy.spatial.distance import pdist
from sklearn.model_selection import train_test_split
import matplotlib.patches as mpatches
from sklearn.manifold import TSNE
from mpl_toolkits.mplot3d import Axes3D
from sklearn.datasets import make_blobs
from sklearn.cluster import KMeans
from itertools import cycle

Dataset_1 = pd.read_csv('dataset_1.csv', index_col=0, low_memory=False)#input data from csv file with the index
Dataset_1 = Dataset_1.T #transform the data -> let features be column name sample be row name
Dataset_1 #check the dataset

#sns.pairplot(Dataset_1)
# 440 feature not choose this to plot all feature for obsercation

#get the data type of the DataFrame content
Dataset_1.dtypes

ACTL6A_S5     object
ADAM9_S2      object
ADAMTS1_S5    object
ADCY7_S3      object
AIMP2_S5      object
               ...  
ZYX_S2        object
MMP3_S2       object
PLG_S5        object
RGS8_S5       object
Subgroup      object
Length: 441, dtype: object

Dataset_1.describe()

#all data are decimal so change the object(string) to float64 -> can have 18 digits after the decimal point
Dataset_1[Dataset_1.columns[0:-1]] = Dataset_1[Dataset_1.columns[0:-1]].astype('float64')

#check the data: never lost data after change data type
Dataset_1

Dataset_1.dtypes

ACTL6A_S5     float64
ADAM9_S2      float64
ADAMTS1_S5    float64
ADCY7_S3      float64
AIMP2_S5      float64
               ...   
ZYX_S2        float64
MMP3_S2       float64
PLG_S5        float64
RGS8_S5       float64
Subgroup       object
Length: 441, dtype: object

#statistic details of the dataset : min max mean std.... 
Dataset_1.describe()

#check the duplicated and show how many
Dataset_1.duplicated(keep=False)
#show the location of the duplicated:
Dataset_1.loc[Dataset_1.duplicated(),:]

#delete duplicated data 
Dataset_1 = Dataset_1.drop_duplicates()

# check the data after drop duplicated data
Dataset_1

sns.histplot(Dataset_1,kde=True, legend=False)
plt.show()

sns.boxplot(data=Dataset_1)#for quartile method

<Axes: >

Dataset_1.describe()

# show the missing data before delet the outliers
plt.figure(figsize=(10,6))
sns.displot(
    data=Dataset_1.isna().melt(value_name="missing"),
    y="variable",
    hue="missing",
    multiple="fill",
    aspect=1.25
)

<seaborn.axisgrid.FacetGrid at 0x29415a6c0d0>

<Figure size 1000x600 with 0 Axes>

#identify function for finding outliers
def find_outliers_IQR(df):
   q1=df.quantile(0.25)#1/4
   q3=df.quantile(0.75)#3/4
   IQR=q3-q1
   df[((df<(q1-1.5*IQR)) | (df>(q3+1.5*IQR)))] =1 #easy to sum how many outliers 
   df[~((df<(q1-1.5*IQR)) | (df>(q3+1.5*IQR)))]=0
   return df

outliers = find_outliers_IQR(Dataset_1[Dataset_1.columns[0:-1]]) #get outliers location
outliers

C:\Users\murui\AppData\Local\Temp\ipykernel_38176\714981418.py:6: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  df[((df<(q1-1.5*IQR)) | (df>(q3+1.5*IQR)))] =1 #easy to sum how many outliers
C:\Users\murui\AppData\Local\Temp\ipykernel_38176\714981418.py:6: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  df[((df<(q1-1.5*IQR)) | (df>(q3+1.5*IQR)))] =1 #easy to sum how many outliers
C:\Users\murui\AppData\Local\Temp\ipykernel_38176\714981418.py:7: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  df[~((df<(q1-1.5*IQR)) | (df>(q3+1.5*IQR)))]=0
C:\Users\murui\AppData\Local\Temp\ipykernel_38176\714981418.py:7: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  df[~((df<(q1-1.5*IQR)) | (df>(q3+1.5*IQR)))]=0

#identify function to drop outliers return a dataset after cleanned
def drop_outliers_IQR(df):
   q1=df.quantile(0.25)#1/4
   q3=df.quantile(0.75)#3/4
   IQR=q3-q1
   not_outliers = df[~((df<(q1-1.5*IQR)) | (df>(q3+1.5*IQR)))] # only get data which is not outliers into dataset
   return not_outliers

cleaned_data = drop_outliers_IQR(Dataset_1[Dataset_1.columns[0:-1]])
cleaned_data

#show the missing data afer cleaning
plt.figure(figsize=(10,6))
sns.displot(
    data=cleaned_data.isna().melt(value_name="missing"),
    y="variable",
    hue="missing",
    multiple="fill",
    aspect=1.25
)

<seaborn.axisgrid.FacetGrid at 0x294158240d0>

<Figure size 1000x600 with 0 Axes>

#compare in the plot
fig, axes = plt.subplots(1,2)
sns.boxplot(data=Dataset_1, ax=axes[0])#orginal data
sns.boxplot(data=cleaned_data, ax=axes[1])#without outliers
axes[0].set_title('never clean')
axes[1].set_title('cleaned')

Text(0.5, 1.0, 'cleaned')

# count the outliers
outliers.sum().sum()

2185.0

#shows situation for count of the outliers in each features
outliers.sum()

ACTL6A_S5      8.0
ADAM9_S2      21.0
ADAMTS1_S5     6.0
ADCY7_S3       1.0
AIMP2_S5       1.0
              ... 
ZWILCH_S5      2.0
ZYX_S2         8.0
MMP3_S2        1.0
PLG_S5        45.0
RGS8_S5        9.0
Length: 440, dtype: float64

#plot for some features for have check 
fig, axes = plt.subplots(1,2)
sns.histplot(Dataset_1[['PLG_S5']],kde=True, legend=False, ax=axes[0])
sns.histplot(cleaned_data[['PLG_S5']],kde=True, legend=False, ax=axes[1])
axes[0].set_title('never clean')
axes[1].set_title('cleaned')
fig, axes = plt.subplots(1,2)
sns.histplot(Dataset_1[['APOC1_S3']],kde=True, legend=False, ax=axes[0])
sns.histplot(cleaned_data[['APOC1_S3']],kde=True, legend=False, ax=axes[1])
axes[0].set_title('never clean')
axes[1].set_title('cleaned')
fig, axes = plt.subplots(1,2)
sns.histplot(Dataset_1[['ACTL6A_S5']],kde=True, legend=False, ax=axes[0])
sns.histplot(cleaned_data[['ACTL6A_S5']],kde=True, legend=False, ax=axes[1])
axes[0].set_title('never clean')
axes[1].set_title('cleaned')
plt.show()

# using isna() to check the missing data and summarise it
Dataset_1.isna().any()

ACTL6A_S5     False
ADAM9_S2      False
ADAMTS1_S5    False
ADCY7_S3      False
AIMP2_S5      False
              ...  
ZYX_S2        False
MMP3_S2       False
PLG_S5        False
RGS8_S5       False
Subgroup      False
Length: 441, dtype: bool

#shows the location of the missing data
Dataset_1[Dataset_1.columns[Dataset_1.isna().any()]]

cleaned_data[cleaned_data.columns[cleaned_data.isna().any()]]

#sum() total of the missing data 
#show count of the missing data in list of features
#Dataset_1.isnull().sum()
#shows total missing data in all dataset -> a number
Dataset_1.isnull().sum().sum()

140

#sum the missing data after cleaning
cleaned_data.isnull().sum().sum()

2326

# sumarize number in each features
Sum_of_missing = Dataset_1.isnull().sum()
print(Sum_of_missing[['CBX1_S5']],'\n',Sum_of_missing[['F3_S5']],'\n',Sum_of_missing[['IGFBP3_S2']],'\n',Sum_of_missing[['IL7R_S3']])

CBX1_S5    0
dtype: int64 
 F3_S5    0
dtype: int64 
 IGFBP3_S2    0
dtype: int64 
 IL7R_S3    0
dtype: int64

# Apply mice imputation (Multiple Imputation by Chained Equation)
mice_imputer = IterativeImputer()

for data_c in Dataset_1.columns[Dataset_1.isna().any()]:
    Dataset_1[[data_c]] = mice_imputer.fit_transform(Dataset_1[[data_c]])

C:\Users\murui\AppData\Local\Temp\ipykernel_38176\3014042716.py:5: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  Dataset_1[[data_c]] = mice_imputer.fit_transform(Dataset_1[[data_c]])
C:\Users\murui\AppData\Local\Temp\ipykernel_38176\3014042716.py:5: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  Dataset_1[[data_c]] = mice_imputer.fit_transform(Dataset_1[[data_c]])
C:\Users\murui\AppData\Local\Temp\ipykernel_38176\3014042716.py:5: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  Dataset_1[[data_c]] = mice_imputer.fit_transform(Dataset_1[[data_c]])
C:\Users\murui\AppData\Local\Temp\ipykernel_38176\3014042716.py:5: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  Dataset_1[[data_c]] = mice_imputer.fit_transform(Dataset_1[[data_c]])

mice_imputer_c = IterativeImputer()

for data_c in cleaned_data.columns[cleaned_data.isna().any()]:
    cleaned_data[[data_c]] = mice_imputer_c.fit_transform(cleaned_data[[data_c]])

Dataset_1.isnull().values.any()

False

cleaned_data.isnull().values.any()

False

Dataset_1.isnull().sum().sum()

0

cleaned_data.isnull().sum().sum()

0

Dataset_1

cleaned_data

data = Dataset_1
# Separate features and labelling
X = data.drop(['Subgroup'], axis=1)  # 440 features
y = data['Subgroup']  # subgroup 

# create Isolation Forest model and training
model = IsolationForest(contamination=0.05, random_state=42)
model.fit(X)

# Predict outliers
y_pred = model.predict(X)

outlier_indices = X.index[y_pred == -1]

# let outliers to NaN
X_cleaned = X.copy()
X_cleaned.loc[outlier_indices] = None


cleaned_data_2 = pd.concat([X_cleaned, y], axis=1)
cleaned_group_labels = cleaned_data_2['Subgroup']
cleaned_data_2 = cleaned_data_2.drop(outlier_indices)

#cleaned_data.to_csv('cleaned_data.csv', index=False)

outlier_indices.value_counts().sum()

27

cleaned_data_2

cleaned_data

cleaned_data_2[cleaned_data_2.columns[0:-1]].max().max()

0.983192666

cleaned_data[cleaned_data.columns[0:-1]].max().max()

0.983192666

cleaned_data_2[cleaned_data_2.columns[0:-1]].min().min()

0.026909471

cleaned_data[cleaned_data.columns[0:-1]].min().min()

0.026909471

#compare in the plot
fig, axes = plt.subplots(1,3)
sns.boxplot(data=Dataset_1, ax=axes[0])
sns.boxplot(data=cleaned_data_2, ax=axes[1])
sns.boxplot(data=cleaned_data, ax=axes[2])
axes[0].set_title('no_clean')
axes[1].set_title('clean_IF')
axes[2].set_title('clean_Q')

Text(0.5, 1.0, 'clean_Q')

#plot for some features for have check 
fig, axes = plt.subplots(1,3)
sns.histplot(Dataset_1[['PLG_S5']],kde=True, legend=False, ax=axes[0])
sns.histplot(cleaned_data_2[['PLG_S5']],kde=True, legend=False, ax=axes[1])
sns.histplot(cleaned_data[['PLG_S5']],kde=True, legend=False, ax=axes[2])
axes[0].set_title('never clean')
axes[1].set_title('clean_IF')
axes[2].set_title('clean_Q')
fig, axes = plt.subplots(1,3)
sns.histplot(Dataset_1[['APOC1_S3']],kde=True, legend=False, ax=axes[0])
sns.histplot(cleaned_data_2[['APOC1_S3']],kde=True, legend=False, ax=axes[1])
sns.histplot(cleaned_data[['APOC1_S3']],kde=True, legend=False, ax=axes[2])
axes[0].set_title('never clean')
axes[1].set_title('clean_IF')
axes[2].set_title('clean_Q')
fig, axes = plt.subplots(1,3)
sns.histplot(Dataset_1[['ACTL6A_S5']],kde=True, legend=False, ax=axes[0])
sns.histplot(cleaned_data_2[['ACTL6A_S5']],kde=True, legend=False, ax=axes[1])
sns.histplot(cleaned_data[['ACTL6A_S5']],kde=True, legend=False, ax=axes[2])
axes[0].set_title('never clean')
axes[1].set_title('clean_IF')
axes[2].set_title('clean_Q')
plt.show()

#formal the data cause the subgroup are object cannot calculate
features = Dataset_1.columns[0:-1]
x = cleaned_data_2.loc[:,features].values
y = Dataset_1.loc[:,['Subgroup']].values

x

array([[0.59234916, 0.62863443, 0.57078532, ..., 0.54840526, 0.14413165,
        0.15669266],
       [0.61670755, 0.58980588, 0.54691471, ..., 0.48897231, 0.12963003,
        0.14766489],
       [0.59158323, 0.60137506, 0.53057323, ..., 0.60270384, 0.23660706,
        0.23660706],
       ...,
       [0.53855485, 0.60062186, 0.5991449 , ..., 0.24772623, 0.17270843,
        0.17270843],
       [0.51361922, 0.59938179, 0.58990848, ..., 0.25990241, 0.1717848 ,
        0.1717848 ],
       [0.54035758, 0.55191784, 0.52615256, ..., 0.47698793, 0.13393862,
        0.13393862]])

#check data type
x.dtype

dtype('float64')

Dataset_max = np.amax(x)
Dataset_max

0.983192666

Dataset_min = np.amin(x)
Dataset_min

0.026909471

sns.histplot(x,kde=True,legend=False)
plt.show()

#Shapiro-Wilk Test: a statistical test that can be used to determine whether a dataset follows a normal distribution.
statistic, p_value = stats.shapiro(x)
if p_value > 0.05:
    print("Data follows a normal distribution")
else:
    print("Data does not follow a normal distribution")

Data does not follow a normal distribution

C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\scipy\stats\_morestats.py:1816: UserWarning: p-value may not be accurate for N > 5000.
  warnings.warn("p-value may not be accurate for N > 5000.")

#apply MinMax
scaler = MinMaxScaler()
X_scaled = scaler.fit_transform(x)

X_scaled

array([[0.46420631, 0.55301778, 0.55043053, ..., 0.67779844, 0.14538856,
        0.22350296],
       [0.5749009 , 0.40554712, 0.48263009, ..., 0.5958461 , 0.12740246,
        0.20524521],
       [0.4607256 , 0.44948682, 0.43621486, ..., 0.75267098, 0.26008414,
        0.38512169],
       ...,
       [0.21974274, 0.44662618, 0.63098116, ..., 0.26319088, 0.18083183,
        0.25589323],
       [0.10642494, 0.4419164 , 0.60474664, ..., 0.27998066, 0.17968626,
        0.25402527],
       [0.22793511, 0.26164852, 0.42365869, ..., 0.57932079, 0.13274633,
        0.17748523]])

np.amax(X_scaled)

1.0000000000000004

np.amin(X_scaled)

0.0

X_scaled.shape

(513, 440)

Dataset_sd = pd.DataFrame(data=X_scaled,
                          columns=cleaned_data_2.columns[0:-1],
                          index=cleaned_data_2.index)
Dataset_sd = pd.concat([Dataset_sd,cleaned_data_2[['Subgroup']]], axis=1)
Dataset_sd

sns.histplot(X_scaled,kde=True,legend=False)
plt.show()

#plot for some features for have check 
fig, axes = plt.subplots(1,2)
sns.histplot(Dataset_sd[['PLG_S5']],kde=True, legend=False, ax=axes[0])
sns.histplot(cleaned_data_2[['PLG_S5']],kde=True, legend=False, ax=axes[1])
axes[0].set_title('Normalisation')
axes[1].set_title('no Normalisation')
fig, axes = plt.subplots(1,2)
sns.histplot(Dataset_sd[['APOC1_S3']],kde=True, legend=False, ax=axes[0])
sns.histplot(cleaned_data_2[['APOC1_S3']],kde=True, legend=False, ax=axes[1])
axes[0].set_title('Normalisation')
axes[1].set_title('no Normalisation')
fig, axes = plt.subplots(1,2)
sns.histplot(Dataset_sd[['ACTL6A_S5']],kde=True, legend=False, ax=axes[0])
sns.histplot(cleaned_data_2[['ACTL6A_S5']],kde=True, legend=False, ax=axes[1])
plt.show()
axes[0].set_title('Normalisation')
axes[1].set_title('no Normalisation')

Text(0.5, 1.0, 'no Normalisation')

# get dataset from standared dataset and format it.
Dataset_nmf = Dataset_sd[Dataset_sd.columns[0:-1]].T 
Dataset_nmf

# Dataset without subgroup column:
X_scaled_nmf = Dataset_nmf[Dataset_nmf.columns[0:-1]]
X_scaled_nmf

#define the data for nmf 
X = X_scaled_nmf
# Define the range of rank (k) values to test
rank_values = range(1, 11)
# Initialize a list to store cophenetic coefficients
cophenetic_coeffs = []

for k in rank_values:
    # Create an NMF model with the current rank (k)
    model = NMF(n_components=k, init='nndsvd', random_state=0) #because all data are positive so don't need to random
    
    # Fit the NMF model to your data
    W = model.fit_transform(X)
    H = model.components_
    
    # Get the linkage matrix for hierarchical clustering
    linkage_matrix = linkage(W, method='ward') 
    
    # Calculate cophenetic coefficient
    coph_coeff, coph_dists = cophenet(linkage_matrix, pdist(X))
    
    # Append the cophenetic coefficient to the list
    cophenetic_coeffs.append(coph_coeff)

# Plot the cophenetic coefficients against rank (k)
plt.plot(rank_values, cophenetic_coeffs, marker='o')
plt.title('Cophenetic Coefficient vs. Rank (k) in NMF')
plt.xlabel('Rank (k)')
plt.ylabel('Cophenetic Coefficient')
plt.grid(True)
plt.show()

C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\sklearn\decomposition\_nmf.py:1710: ConvergenceWarning: Maximum number of iterations 200 reached. Increase it to improve convergence.
  warnings.warn(
C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\sklearn\decomposition\_nmf.py:1710: ConvergenceWarning: Maximum number of iterations 200 reached. Increase it to improve convergence.
  warnings.warn(
C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\sklearn\decomposition\_nmf.py:1710: ConvergenceWarning: Maximum number of iterations 200 reached. Increase it to improve convergence.
  warnings.warn(
C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\sklearn\decomposition\_nmf.py:1710: ConvergenceWarning: Maximum number of iterations 200 reached. Increase it to improve convergence.
  warnings.warn(
C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\sklearn\decomposition\_nmf.py:1710: ConvergenceWarning: Maximum number of iterations 200 reached. Increase it to improve convergence.
  warnings.warn(
C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\sklearn\decomposition\_nmf.py:1710: ConvergenceWarning: Maximum number of iterations 200 reached. Increase it to improve convergence.
  warnings.warn(
C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\sklearn\decomposition\_nmf.py:1710: ConvergenceWarning: Maximum number of iterations 200 reached. Increase it to improve convergence.
  warnings.warn(
C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\sklearn\decomposition\_nmf.py:1710: ConvergenceWarning: Maximum number of iterations 200 reached. Increase it to improve convergence.
  warnings.warn(
C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\sklearn\decomposition\_nmf.py:1710: ConvergenceWarning: Maximum number of iterations 200 reached. Increase it to improve convergence.
  warnings.warn(
C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\sklearn\decomposition\_nmf.py:1710: ConvergenceWarning: Maximum number of iterations 200 reached. Increase it to improve convergence.
  warnings.warn(

n_components_1 = 3
n_components_2 = 10

#fit and transform the machine learning model by dataset with two values for parameter n_components
model = NMF(n_components_1, init='nndsvd', random_state=0)#5 is the best rank...
W = model.fit_transform(Dataset_nmf) #nmf features
H = model.components_

C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\sklearn\decomposition\_nmf.py:1710: ConvergenceWarning: Maximum number of iterations 200 reached. Increase it to improve convergence.
  warnings.warn(

model2 = NMF(n_components_2, init='nndsvd', random_state=0)#5 is the best rank...
W2 = model2.fit_transform(Dataset_nmf) #nmf features
H2 = model2.components_

C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\sklearn\decomposition\_nmf.py:1710: ConvergenceWarning: Maximum number of iterations 200 reached. Increase it to improve convergence.
  warnings.warn(

print('error_value for rank = 3: ', model.reconstruction_err_, ' vs ', 'error_value for rank = 10: ', model2.reconstruction_err_)

error_value for rank = 3:  63.920348014243686  vs  error_value for rank = 10:  51.31911989982033

H.shape

(3, 513)

compontents_df = pd.DataFrame(H, 
                              columns=Dataset_nmf.columns)
compontents_df_2 = compontents_df.T
compontents_df_2 = pd.concat([compontents_df.T,Dataset_sd[['Subgroup']]],axis=1)
compontents_df_2 = compontents_df_2.T
compontents_df

compontents_df_3 = pd.DataFrame(H2, 
                              columns=Dataset_nmf.columns)
compontents_df_4 = compontents_df_3.T
compontents_df_4 = pd.concat([compontents_df_3.T,Dataset_sd[['Subgroup']]],axis=1)
compontents_df_4 = compontents_df_4.T
compontents_df_3

W.shape

(440, 3)

compontents_df_W = pd.DataFrame(W,
                                index=Dataset_nmf.index)
compontents_df_W = compontents_df_W.T
compontents_df_W

compontents_df_W2 = pd.DataFrame(W2,
                                index=Dataset_nmf.index)
compontents_df_W2 = compontents_df_W2.T
compontents_df_W2

sns.set(font_scale=1)  # Adjust the font size if needed
sns.clustermap(compontents_df, cmap='viridis', method='average', metric='euclidean', figsize=(8, 6))
plt.title('coefmap in n_component = 3')
# Show the plot
plt.show()

sns.set(font_scale=1)  # Adjust the font size if needed
sns.clustermap(compontents_df_W, cmap='viridis', method='average', metric='euclidean', figsize=(8, 6))
plt.title('basismap in n_component = 3')
# Show the plot
plt.show()

sns.set(font_scale=1)  # Adjust the font size if needed
sns.clustermap(compontents_df_3, cmap='viridis', method='average', metric='euclidean', figsize=(8, 6))
plt.title('coefmap in component = 10')
# Show the plot
plt.show()

sns.set(font_scale=1)  # Adjust the font size if needed
sns.clustermap(compontents_df_W2, cmap='viridis', method='average', metric='euclidean', figsize=(8, 6))
plt.title('basismap in n_component = 10')
# Show the plot
plt.show()

# Create a sample dataset (replace this with your own data)
data = np.dot(W, H)

# Create a DataFrame for your NMF data
df = pd.DataFrame(data, columns=range(data.shape[1]))

# Create a clustermap
sns.clustermap(df, cmap='coolwarm', method='ward', figsize=(8, 6))
plt.title('rank = 10')
plt.xlabel("W")
plt.ylabel("H")
# Display the plot
plt.show()

C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\seaborn\matrix.py:560: UserWarning: Clustering large matrix with scipy. Installing `fastcluster` may give better performance.
  warnings.warn(msg)
C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\seaborn\matrix.py:560: UserWarning: Clustering large matrix with scipy. Installing `fastcluster` may give better performance.
  warnings.warn(msg)

# Create a sample dataset (replace this with your own data)
data = np.dot(W2, H2)

# Create a DataFrame for your NMF data
df = pd.DataFrame(data, columns=range(data.shape[1]))

# Create a clustermap
sns.clustermap(df, cmap='coolwarm', method='ward', figsize=(8, 6))
plt.title('rank = 10')
plt.xlabel("W")
plt.ylabel("H")
# Display the plot
plt.show()

C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\seaborn\matrix.py:560: UserWarning: Clustering large matrix with scipy. Installing `fastcluster` may give better performance.
  warnings.warn(msg)
C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\seaborn\matrix.py:560: UserWarning: Clustering large matrix with scipy. Installing `fastcluster` may give better performance.
  warnings.warn(msg)

#for feature analysis
plt.figure(figsize=(8, 6))
plt.imshow(W, cmap='viridis', aspect='auto')
plt.colorbar()
plt.title("Matrix W Heatmap in n_component = 3")
plt.xlabel("rank k")
plt.ylabel("Features")
plt.show()

X = X_scaled_nmf
# Apply NMF for dimensionality reduction
nmf = NMF(n_components=2)
X_nmf = nmf.fit_transform(X)

# Perform K-Means clustering into 6 clusters -> subgroup are 6
kmeans = KMeans(n_clusters=6)
labels = kmeans.fit_predict(X_nmf)

# Create a list of colors for plotting
colors = cycle(['b', 'g', 'r', 'c', 'm', 'y'])  # You can extend this list for more colors

# Visualize the scatter plot with different colors for each cluster
plt.figure(figsize=(8, 6))
for cluster_num, color in zip(range(6), colors):
    cluster_points = X_nmf[labels == cluster_num]
    plt.scatter(cluster_points[:, 0], cluster_points[:, 1,], c=color, label=f'Cluster {cluster_num}')

# Label clusters
for cluster in range(6):
    cluster_center = X_nmf[labels == cluster].mean(axis=0)
    plt.text(cluster_center[0], cluster_center[1], f'Cluster {cluster}', fontsize=12)

plt.title("NMF Scatter Plot with Cluster Labels")
plt.legend()
plt.show()

C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\sklearn\decomposition\_nmf.py:1710: ConvergenceWarning: Maximum number of iterations 200 reached. Increase it to improve convergence.
  warnings.warn(
C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\sklearn\cluster\_kmeans.py:1416: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning
  super()._check_params_vs_input(X, default_n_init=10)

# Generate some example data
X= X_scaled_nmf

# Apply NMF for dimensionality reduction
nmf = NMF(n_components=3)  # Use 3 components for 3D
X_nmf = nmf.fit_transform(X)

# Perform K-Means clustering into 6 clusters
kmeans = KMeans(n_clusters=6)
labels = kmeans.fit_predict(X_nmf)

# Create a list of colors for plotting
colors = cycle('bgrcmy')  # You can extend this for more colors

# Visualize the 3D scatter plot with different colors for each cluster
fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')

for cluster_num, color in zip(range(6), colors):
    cluster_points = X_nmf[labels == cluster_num]
    ax.scatter(cluster_points[:, 0], cluster_points[:, 1], cluster_points[:, 2], c=color, label=f'Cluster {cluster_num}')

ax.set_title("3D Scatter Plot with 6 Clusters")
ax.legend(loc='best')
plt.show()

C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\sklearn\decomposition\_nmf.py:1710: ConvergenceWarning: Maximum number of iterations 200 reached. Increase it to improve convergence.
  warnings.warn(
C:\Users\murui\AppData\Roaming\Python\Python311\site-packages\sklearn\cluster\_kmeans.py:1416: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning
  super()._check_params_vs_input(X, default_n_init=10)

#format dataset
features = Dataset_1.columns[0:-1]
x_pca = Dataset_sd.loc[:,features].values
y_pca = Dataset_sd.loc[:,['Subgroup']].values

pca = PCA(n_components=0.9) # set pca function for 90% data
Dataset_pca = pca.fit_transform(x_pca)
Dataset_pca

array([[-0.22086956, -1.57353236,  1.62778104, ...,  0.18280888,
        -0.03627199,  0.06702992],
       [-0.65072698, -0.77176889,  0.46795313, ...,  0.18633462,
         0.02753773,  0.15819418],
       [ 0.82876425, -1.7883118 , -0.24780858, ..., -0.35957439,
         0.04768204, -0.36755787],
       ...,
       [-1.30179571, -0.07552246, -0.13932056, ...,  0.07359615,
         0.14471046, -0.03762028],
       [-2.10909455,  1.16512941, -0.79659035, ...,  0.00502626,
        -0.27785708,  0.15322508],
       [-1.86684625,  0.39593328, -1.37928861, ..., -0.06436798,
         0.06116171,  0.04452186]])

print(pca.explained_variance_ratio_.cumsum())

[0.23706203 0.40262552 0.45346892 0.49663741 0.5327645  0.56335521
 0.5873575  0.60933684 0.62712978 0.6432445  0.65793841 0.67151867
 0.68296563 0.69411493 0.70398508 0.71266598 0.72048057 0.72786919
 0.73464777 0.74094081 0.74713051 0.75312785 0.75881736 0.76415767
 0.76932186 0.77424919 0.77886561 0.78339976 0.78756513 0.79163156
 0.79563607 0.79937535 0.80293863 0.80638147 0.80973273 0.81303144
 0.81622553 0.81928301 0.82220704 0.82493434 0.82765238 0.83032314
 0.83289089 0.83536028 0.8378268  0.84023843 0.84264276 0.84497953
 0.8472294  0.84941442 0.85156252 0.8536918  0.85574011 0.85777526
 0.85975976 0.86171791 0.86358299 0.86543353 0.86725918 0.86904035
 0.87081097 0.87257487 0.87427877 0.87591737 0.8775382  0.87913788
 0.88073018 0.88229619 0.88383964 0.88537677 0.88684614 0.88830666
 0.88975582 0.89118181 0.89257139 0.89394927 0.89531181 0.89663958
 0.89795251 0.89925148 0.90052438]

pca1 = PCA()
pca1.fit_transform(x_pca)
plt.plot(pca1.explained_variance_ratio_.cumsum(), marker='o', linestyle='-')#plot variance_sum follow the principal
plt.axhline(y=0.9, color='b', linestyle='--') # plot a line when y=1 cause this is when principal include 90% data
plt.text(0.5, 0.92, '90% cut-off threshold', color = 'blue', fontsize=10)
plt.axhline(y=0.7, color='g', linestyle='--') # plot a line when y=1 cause this is when principal include 70% data
plt.text(0.5, 0.72, '70% cut-off threshold', color = 'green', fontsize=10)
plt.axhline(y=0.5, color='purple', linestyle='--') # plot a line when y=1 cause this is when principal include 50% data
plt.text(0.5, 0.52, '50% cut-off threshold', color = 'purple', fontsize=10)
plt.title('Explained Variance by Components')
plt.xlabel("The Number of Components")
plt.ylabel("Cumulative Explained Variances" )
plt.show()

plt.hist([round(a, 4) for a in pca.explained_variance_ratio_], density=False, color = 'lightblue', cumulative = False, label = "V")
plt.legend()
plt.show()

df_pca_components = pca1.explained_variance_ratio_.cumsum() > 0.90
Number_of_components_percentage =  440 - df_pca_components.sum()
Number_of_components_percentage

80

Dataset_pca = pd.DataFrame(data=Dataset_pca.T[0:3].T,
                            columns=['PC1','PC2','PC3'],
                            index=Dataset_sd.index)
Dataset_pca

Dataset_pca = pd.concat([Dataset_pca,Dataset_sd[['Subgroup']]], axis=1)
Dataset_pca

# plot 3D:
fig = plt.figure(figsize=(8,8))
ax = plt.axes(projection='3d')

# Define colours
color_map = {'C1': 'red', 'C2': 'yellow', 'C3':'green', 'C4':'cyan', 'C5': 'blue', 'C6': 'purple'}

# Data for three-dimensional scattered points
zdata = Dataset_pca['PC1']
xdata = Dataset_pca['PC2']
ydata = Dataset_pca['PC3']

ax.set_xlabel('PC1')
ax.set_ylabel('PC2')
ax.set_zlabel('PC3')

ax.view_init(-165, 120)

ax.scatter3D(xdata, ydata, zdata, c=Dataset_pca['Subgroup'].map(color_map))

c1 = mpatches.Patch(color='red', label='C1')
c2 = mpatches.Patch(color='yellow', label='C2')
c3 = mpatches.Patch(color='green', label='C3')
c4 = mpatches.Patch(color='cyan', label='C4')
c5 = mpatches.Patch(color='blue', label='C5')
c6 = mpatches.Patch(color='purple', label='C6')

plt.legend(handles=[c1, c2, c3, c4, c5, c6])

<matplotlib.legend.Legend at 0x29453b0c1d0>

fig = plt.figure(figsize=(8,8))
ax = plt.axes(projection='3d')

# Define colours
color_map = {'C1': 'red', 'C2': 'yellow', 'C3':'green', 'C4':'cyan', 'C5': 'blue', 'C6': 'purple'}

# Data for three-dimensional scattered points
zdata = Dataset_pca['PC1']
xdata = Dataset_pca['PC2']
ydata = Dataset_pca['PC3']

ax.set_xlabel('PC1')
ax.set_ylabel('PC2')
ax.set_zlabel('PC3')

ax.view_init(-160, 200)

ax.scatter3D(xdata, ydata, zdata, c=Dataset_pca['Subgroup'].map(color_map))

c1 = mpatches.Patch(color='red', label='C1')
c2 = mpatches.Patch(color='yellow', label='C2')
c3 = mpatches.Patch(color='green', label='C3')
c4 = mpatches.Patch(color='cyan', label='C4')
c5 = mpatches.Patch(color='blue', label='C5')
c6 = mpatches.Patch(color='purple', label='C6')

plt.legend(handles=[c1, c2, c3, c4, c5, c6])

<matplotlib.legend.Legend at 0x29453b75490>

features = Dataset_1.columns[0:-1]
x_tsne = Dataset_sd.loc[:,features].values
y_tsne = Dataset_sd.loc[:,['Subgroup']].values

X = x_tsne 


# Initialise t-SNE model, set different n_components values and visualise
n_components_list = [2, 3]  # 
perplexity = 30 

for n_components in n_components_list:
    tsne = TSNE(n_components=n_components, perplexity=perplexity)
    X_tsne = tsne.fit_transform(X)

    # 可视化降维后的数据
    if n_components == 2:
        plt.figure(figsize=(6, 6))
        plt.scatter(X_tsne[:, 0], X_tsne[:, 1])
        plt.title(f't-SNE (n_components={n_components})')
        plt.show()
    elif n_components == 3:
        
        fig = plt.figure()
        ax = fig.add_subplot(111, projection='3d')
        ax.scatter(X_tsne[:, 0], X_tsne[:, 1], X_tsne[:, 2])
        ax.set_title(f't-SNE (n_components={n_components})')
        plt.show()

tsne = TSNE(n_components=3, random_state=0)
X_3d = tsne.fit_transform(x_tsne)

tsne_1 = TSNE(n_components=2, random_state=0)
X_2d = tsne_1.fit_transform(x_tsne)

X_3d

array([[ -1.6142616, -11.426699 ,  -5.145396 ],
       [ -1.0413069,  -1.5464859,  -3.6518073],
       [  4.791106 ,  -4.6398287,  -1.5009079],
       ...,
       [ -5.3804784,   0.6474102,   3.2778208],
       [ -8.747445 ,   2.0955217,  -5.7901073],
       [ -4.704082 ,  -1.6342709,  -2.8095233]], dtype=float32)

Dataset_tsne = pd.DataFrame(data = X_3d,
                            columns=['feature1','feature2','feature3'],
                            index=Dataset_sd.index)
Dataset_tsne

Dataset_tsne1 = pd.DataFrame(data = X_2d,
                            columns=['feature1','feature2'],
                            index=Dataset_sd.index)
Dataset_tsne1

Dataset_tsne = pd.concat([Dataset_tsne,Dataset_sd[['Subgroup']]], axis=1)
Dataset_tsne

Dataset_tsne1 = pd.concat([Dataset_tsne1,Dataset_sd[['Subgroup']]], axis=1)
Dataset_tsne1

fig = plt.figure(figsize=(8,8))
ax = fig.add_subplot(111, projection='3d')

ax.scatter(X_3d[:, 0], X_3d[:, 1], X_3d[:, 2], c='b', marker='o')

ax.set_xlabel('t-SNE Dimension 1')
ax.set_ylabel('t-SNE Dimension 2')
ax.set_zlabel('t-SNE Dimension 3')
ax.view_init(-160, 220)
targets = ['C1', 'C2', 'C3','C4','C5','C6']
colors = ['red', 'yellow', 'green','cyan','blue','purple']
for target, color in zip(targets,colors):
    indicesToKeep = Dataset_tsne['Subgroup'] == target
    ax.scatter(Dataset_tsne.loc[indicesToKeep, 'feature1']
               , Dataset_tsne.loc[indicesToKeep, 'feature2']
               , Dataset_tsne.loc[indicesToKeep, 'feature3']
               , c = color
               , s = 50)
ax.legend(targets)
ax.grid()

fig = plt.figure(figsize=(8,8))
ax = fig.add_subplot(111, projection='3d')

ax.scatter(X_3d[:, 0], X_3d[:, 1], X_3d[:, 2], c='b', marker='o')

ax.set_xlabel('t-SNE Dimension 1')
ax.set_ylabel('t-SNE Dimension 2')
ax.set_zlabel('t-SNE Dimension 3')
ax.view_init(-160, 200)
targets = ['C1', 'C2', 'C3','C4','C5','C6']
colors = ['red', 'yellow', 'green','cyan','blue','purple']
for target, color in zip(targets,colors):
    indicesToKeep = Dataset_tsne['Subgroup'] == target
    ax.scatter(Dataset_tsne.loc[indicesToKeep, 'feature1']
               , Dataset_tsne.loc[indicesToKeep, 'feature2']
               , Dataset_tsne.loc[indicesToKeep, 'feature3']
               , c = color
               , s = 50)
ax.legend(targets)
ax.grid()

fig = plt.figure(figsize = (8,8)) #size of the plot
ax = fig.add_subplot(1,1,1) #Add an Axes to the current figure or retrieve an existing Axes.
ax.set_xlabel('t-SNE Dimension 1', fontsize = 15)
ax.set_ylabel('t-SNE Dimension 2', fontsize = 15)
ax.set_title('2 Dimension t-SNE', fontsize=20)
targets = ['C1', 'C2', 'C3','C4','C5','C6']
colors = ['red', 'yellow', 'green','cyan','blue','purple']
for target, color in zip(targets,colors):
    indicesToKeep = Dataset_tsne1['Subgroup'] == target
    ax.scatter(Dataset_tsne1.loc[indicesToKeep, 'feature1']
               , Dataset_tsne1.loc[indicesToKeep, 'feature2']
               , c = color
               , s = 50)
ax.legend(targets)#group point by targets
ax.grid()

	ACTL6A_S5	ADAM9_S2	ADAMTS1_S5	ADCY7_S3	AIMP2_S5	ALKBH7_S5	ALOX5AP_S3	AMPD3_S3	APITD1_S5	APOC1_S3	...	WIPF1_S3	WNT2B_S2	WNT8B_S2	WSB2_S5	ZWILCH_S5	ZYX_S2	MMP3_S2	PLG_S5	RGS8_S5	Subgroup
TCGA.HF.7136.STAD.C1_	0.592349159	0.628634428	0.570785319	0.536637178	0.509823444	0.462217686	0.443087096	0.482923423	0.418518998	0.483556486	...	0.557670757	0.348505759	0.21001634	0.629422932	0.515652808	0.664830576	0.548405265	0.144131649	0.156692658	C1
TCGA.EW.A1OZ.BRCA.C1_	0.616707551	0.589805879	0.546914715	0.522103124	0.523892866	0.545415692	0.518862114	0.428246143	0.447287266	0.588167354	...	0.543130808	0.271636495	0.215561609	0.588503202	0.520021651	0.661330832	0.48897231	0.129630028	0.147664886	C1
TCGA.EY.A1GL.UCEC.C1_	0.591583228	0.60137506	0.530573235	0.470163893	0.573793291	0.560062178	0.408029142	0.529846796	0.516928306	0.500290877	...	0.537821188	0.370267472	0.256228773	0.626341995	0.500481153	0.650542802	0.602703844	0.236607056	0.236607056	C1
TCGA.EB.A5FP.SKCM.C1_	0.60418886	0.586197969	0.566643527	0.523075884	0.575625221	0.497917854	0.362421719	0.428469267	0.544391175	0.580779692	...	0.635673186	0.360680436	0.261152768	0.612561409	0.564330771	0.620690208	0.457935414	0.261152768	0.261152768	C1
TCGA.D1.A16D.UCEC.C1_	0.545902837	0.632497249	0.58791722	0.533485849	0.540957612	0.556670227	0.518468063	0.510708281	0.488153784	0.585869525	...	0.533509926	0.332905555	0.129105571	0.60903958	0.482094262	0.670960037	0.38913796	0.282478711	0.151171985	C1
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
TCGA.FB.AAQ1.PAAD.C6_.2	0.547088474	0.669353783	0.556373656	0.575362692	0.511424709	0.534338128	0.584272643	0.474198022	0.450351433	0.575779881	...	0.548371597	0.301071951	0.202440846	0.554636793	0.47268884	0.653400943	0.307704021	0.332027894	0.125693204	C6
TCGA.DX.A1KU.SARC.C6_.1	0.54993579	0.652694922	0.674654449	0.521712202	0.510103123	0.50135204	0.521481461	0.501617299	0.430389962	0.517204617	...	0.63649641	0.431338453	0.18709252	0.586367439	0.486879106	0.666805202	0.485531743	0.163455699	0.216324944	C6
TCGA.A2.A4S1.BRCA.C6_	0.513619217	0.599381792	0.589908476	0.543181612	0.4788955	0.507401046	0.570468374	0.483157655	0.441636966	0.46383583	...	0.601900974	0.340523965	0.171784796	0.541963837	0.450655279	0.643585011	0.25990241	0.171784796	0.171784796	C6
TCGA.3U.A98I.MESO.C6_.1	0.504295371	0.620086907	0.602403486	0.542429285	0.476758106	0.503558058	0.511588826	0.457273386	0.436591435	0.539356545	...	0.585780294	0.476789408	0.182592379	0.563271127	0.493495531	0.659346701	0.204456653	0.146536279	0.182592379	C6
TCGA.AR.A2LJ.BRCA.C6_	0.540357583	0.55191784	0.526152563	0.489544812	0.518348041	0.603053115	0.517177218	0.453268531	0.484431339	0.650399976	...	0.563844585	0.215692689	0.201434471	0.558990676	0.462621113	0.676615381	0.476987934	0.133938618	0.133938618	C6

	ACTL6A_S5	ADAM9_S2	ADAMTS1_S5	ADCY7_S3	AIMP2_S5	ALKBH7_S5	ALOX5AP_S3	AMPD3_S3	APITD1_S5	APOC1_S3	...	WIPF1_S3	WNT2B_S2	WNT8B_S2	WSB2_S5	ZWILCH_S5	ZYX_S2	MMP3_S2	PLG_S5	RGS8_S5	Subgroup
TCGA.HF.7136.STAD.C1_	0.592349	0.628634	0.570785	0.536637	0.509823	0.462218	0.443087	0.482923	0.418519	0.483556	...	0.557671	0.348506	0.210016	0.629423	0.515653	0.664831	0.548405	0.144132	0.156693	C1
TCGA.EW.A1OZ.BRCA.C1_	0.616708	0.589806	0.546915	0.522103	0.523893	0.545416	0.518862	0.428246	0.447287	0.588167	...	0.543131	0.271636	0.215562	0.588503	0.520022	0.661331	0.488972	0.129630	0.147665	C1
TCGA.EY.A1GL.UCEC.C1_	0.591583	0.601375	0.530573	0.470164	0.573793	0.560062	0.408029	0.529847	0.516928	0.500291	...	0.537821	0.370267	0.256229	0.626342	0.500481	0.650543	0.602704	0.236607	0.236607	C1
TCGA.EB.A5FP.SKCM.C1_	0.604189	0.586198	0.566644	0.523076	0.575625	0.497918	0.362422	0.428469	0.544391	0.580780	...	0.635673	0.360680	0.261153	0.612561	0.564331	0.620690	0.457935	0.261153	0.261153	C1
TCGA.D1.A16D.UCEC.C1_	0.545903	0.632497	0.587917	0.533486	0.540958	0.556670	0.518468	0.510708	0.488154	0.585870	...	0.533510	0.332906	0.129106	0.609040	0.482094	0.670960	0.389138	0.282479	0.151172	C1
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
TCGA.FB.AAQ1.PAAD.C6_.2	0.547088	0.669354	0.556374	0.575363	0.511425	0.534338	0.584273	0.474198	0.450351	0.575780	...	0.548372	0.301072	0.202441	0.554637	0.472689	0.653401	0.307704	0.332028	0.125693	C6
TCGA.DX.A1KU.SARC.C6_.1	0.549936	0.652695	0.674654	0.521712	0.510103	0.501352	0.521481	0.501617	0.430390	0.517205	...	0.636496	0.431338	0.187093	0.586367	0.486879	0.666805	0.485532	0.163456	0.216325	C6
TCGA.A2.A4S1.BRCA.C6_	0.513619	0.599382	0.589908	0.543182	0.478895	0.507401	0.570468	0.483158	0.441637	0.463836	...	0.601901	0.340524	0.171785	0.541964	0.450655	0.643585	0.259902	0.171785	0.171785	C6
TCGA.3U.A98I.MESO.C6_.1	0.504295	0.620087	0.602403	0.542429	0.476758	0.503558	0.511589	0.457273	0.436591	0.539357	...	0.585780	0.476789	0.182592	0.563271	0.493496	0.659347	0.204457	0.146536	0.182592	C6
TCGA.AR.A2LJ.BRCA.C6_	0.540358	0.551918	0.526153	0.489545	0.518348	0.603053	0.517177	0.453269	0.484431	0.650400	...	0.563845	0.215693	0.201434	0.558991	0.462621	0.676615	0.476988	0.133939	0.133939	C6

	ACTL6A_S5	ADAM9_S2	ADAMTS1_S5	ADCY7_S3	AIMP2_S5	ALKBH7_S5	ALOX5AP_S3	AMPD3_S3	APITD1_S5	APOC1_S3	...	WDR77_S5	WIPF1_S3	WNT2B_S2	WNT8B_S2	WSB2_S5	ZWILCH_S5	ZYX_S2	MMP3_S2	PLG_S5	RGS8_S5
count	588.000000	588.000000	588.000000	588.000000	588.000000	588.000000	588.000000	588.000000	588.000000	588.000000	...	588.000000	588.000000	588.000000	588.000000	588.000000	588.000000	588.000000	588.000000	588.000000	588.000000
mean	0.561551	0.617498	0.548594	0.530525	0.531586	0.550029	0.483884	0.493659	0.477836	0.573563	...	0.550060	0.568496	0.355256	0.210675	0.607736	0.494147	0.643149	0.321869	0.219206	0.192180
std	0.031263	0.045431	0.057663	0.039461	0.037694	0.045562	0.053663	0.044309	0.036428	0.086896	...	0.031076	0.036815	0.074243	0.063276	0.034675	0.035358	0.026195	0.130251	0.147881	0.078831
min	0.454512	0.481747	0.367899	0.420165	0.394824	0.406076	0.331671	0.368469	0.322530	0.324714	...	0.469474	0.445756	0.175897	0.055078	0.491753	0.393490	0.542554	0.056857	0.026909	0.046179
25%	0.540899	0.595816	0.508768	0.502042	0.504046	0.519404	0.445604	0.465907	0.452058	0.517147	...	0.526566	0.543102	0.303990	0.165478	0.581120	0.468319	0.626829	0.220870	0.137597	0.134364
50%	0.558783	0.617728	0.544527	0.532039	0.535152	0.551989	0.486738	0.492044	0.477274	0.574435	...	0.551257	0.572505	0.349658	0.208920	0.603647	0.492761	0.640914	0.287781	0.186938	0.178872
75%	0.581267	0.645072	0.586353	0.559252	0.559505	0.581758	0.521481	0.516225	0.502244	0.620487	...	0.571871	0.596134	0.397857	0.253542	0.634212	0.518996	0.660907	0.420708	0.232155	0.239960
max	0.710251	0.746323	0.729066	0.678374	0.636284	0.679299	0.616650	0.682606	0.569738	0.908518	...	0.648555	0.666154	0.611041	0.455048	0.704738	0.613592	0.745479	0.782070	0.833178	0.540641

	ACTL6A_S5	ADAM9_S2	ADAMTS1_S5	ADCY7_S3	AIMP2_S5	ALKBH7_S5	ALOX5AP_S3	AMPD3_S3	APITD1_S5	APOC1_S3	...	WIPF1_S3	WNT2B_S2	WNT8B_S2	WSB2_S5	ZWILCH_S5	ZYX_S2	MMP3_S2	PLG_S5	RGS8_S5	Subgroup
TCGA.EB.A5FP.SKCM.C1_.1	0.604189	0.586198	0.566644	0.523076	0.575625	0.497918	0.362422	0.428469	0.544391	0.580780	...	0.635673	0.360680	0.261153	0.612561	0.564331	0.620690	0.457935	0.261153	0.261153	C1
TCGA.31.1944.OV.C1_.1	0.603431	0.607367	0.558262	0.514740	0.511529	0.553733	0.495510	0.474454	0.486775	0.533041	...	0.515596	0.348182	0.146301	0.605923	0.500313	0.691955	0.142887	0.741099	0.130762	C1
TCGA.DK.A1A7.BLCA.C1_.1	0.610827	0.547774	0.588034	0.463769	0.499116	0.570002	0.399470	0.465571	0.523996	0.528008	...	0.546360	0.345150	0.235206	0.636773	0.535029	0.616241	0.592075	0.161456	0.161456	C1
TCGA.EW.A1OZ.BRCA.C1_.1	0.616708	0.589806	0.546915	0.522103	0.523893	0.545416	0.518862	0.428246	0.447287	0.588167	...	0.543131	0.271636	0.215562	0.588503	0.520022	0.661331	0.488972	0.129630	0.147665	C1
TCGA.E2.A15M.BRCA.C2_.1	0.558220	0.635687	0.545919	0.547288	0.459488	0.418649	0.512531	0.467458	0.416029	0.569235	...	0.621682	0.278447	0.074778	0.593841	0.505676	0.581101	0.458070	0.061834	0.061834	C2
TCGA.25.2396.OV.C2_.1	0.602502	0.561905	0.607953	0.536461	0.529014	0.589127	0.468006	0.472193	0.510102	0.517969	...	0.543178	0.311914	0.169489	0.570444	0.496169	0.660765	0.182190	0.137118	0.174494	C2
TCGA.FY.A3WA.THCA.C3_.1	0.576086	0.650131	0.634087	0.490138	0.543803	0.509902	0.429367	0.496674	0.549276	0.519384	...	0.581558	0.346849	0.242005	0.677258	0.488890	0.656424	0.227736	0.227736	0.503648	C3
TCGA.5L.AAT0.BRCA.C3_.1	0.556582	0.530522	0.480498	0.525609	0.498281	0.560936	0.524850	0.484109	0.466545	0.624061	...	0.582457	0.315964	0.169491	0.581383	0.480981	0.645153	0.550348	0.103449	0.133479	C3
TCGA.EM.A2CO.THCA.C3_.1	0.566019	0.644224	0.604911	0.479800	0.619699	0.620932	0.469490	0.501437	0.555862	0.467601	...	0.557036	0.307937	0.231493	0.673924	0.479143	0.658761	0.251717	0.231493	0.251717	C3
TCGA.G9.7509.PRAD.C3_.1	0.581532	0.668286	0.605955	0.453622	0.552538	0.615623	0.411858	0.572722	0.527072	0.425430	...	0.544120	0.424376	0.340392	0.628670	0.519053	0.629964	0.254507	0.274949	0.221177	C3
TCGA.BK.A4ZD.UCEC.C4_.1	0.599205	0.661167	0.566843	0.563508	0.543773	0.539424	0.406967	0.398848	0.479347	0.556737	...	0.495199	0.374123	0.275463	0.644563	0.526226	0.673062	0.400839	0.230479	0.230479	C4
TCGA.26.5134.GBM.C4_.1	0.564803	0.604823	0.528689	0.535618	0.507838	0.529862	0.497769	0.487410	0.486818	0.636640	...	0.603814	0.342861	0.298492	0.636048	0.533901	0.654626	0.153145	0.153145	0.153145	C4
TCGA.DD.AACX.LIHC.C4_.1	0.550832	0.602355	0.527694	0.491883	0.569211	0.600703	0.451288	0.419541	0.504886	0.823788	...	0.491883	0.302988	0.326748	0.633527	0.521093	0.618170	0.254250	0.754658	0.254250	C4
TCGA.4G.AAZO.CHOL.C4_.1	0.564792	0.654620	0.499157	0.527904	0.527171	0.572092	0.485063	0.523490	0.464568	0.605643	...	0.555055	0.399375	0.233583	0.590268	0.521786	0.634882	0.199747	0.199747	0.199747	C4
TCGA.CZ.5454.KIRC.C4_.1	0.538890	0.648853	0.588471	0.515182	0.457138	0.511039	0.500226	0.497881	0.446098	0.703095	...	0.602546	0.232489	0.145300	0.579804	0.441574	0.613291	0.215119	0.419721	0.145300	C4
TCGA.E1.A7YM.LGG.C5_.1	0.543830	0.603269	0.534986	0.534850	0.564547	0.565591	0.541178	0.553608	0.472036	0.609623	...	0.610589	0.384765	0.265180	0.636485	0.493780	0.637238	0.205295	0.205295	0.255208	C5
TCGA.DU.5871.LGG.C5_.1	0.563364	0.625737	0.462978	0.576733	0.549523	0.570429	0.528729	0.536600	0.471191	0.609544	...	0.599033	0.390630	0.309956	0.645211	0.445841	0.616982	0.239331	0.221897	0.271622	C5
TCGA.F6.A8O3.LGG.C5_.1	0.574676	0.648779	0.494008	0.591905	0.595162	0.596090	0.437184	0.460764	0.468736	0.556510	...	0.598023	0.381721	0.258388	0.657806	0.509657	0.667191	0.258388	0.245990	0.352207	C5
TCGA.S9.A6WE.LGG.C5_.1	0.559885	0.622396	0.554979	0.552939	0.558326	0.580886	0.485884	0.478218	0.459763	0.632354	...	0.613009	0.392358	0.250723	0.641029	0.434459	0.640533	0.285710	0.230978	0.265094	C5
TCGA.HW.7486.LGG.C5_.1	0.549327	0.626137	0.539728	0.555930	0.564257	0.557118	0.498693	0.587979	0.447479	0.643792	...	0.635534	0.363992	0.253542	0.655838	0.466008	0.616956	0.234051	0.203934	0.307448	C5
TCGA.P5.A5EV.LGG.C5_.1	0.548458	0.611815	0.444916	0.585870	0.567410	0.585895	0.487849	0.502463	0.509270	0.591073	...	0.603650	0.448229	0.208168	0.659484	0.446377	0.631684	0.208168	0.208168	0.226906	C5
TCGA.WY.A85D.LGG.C5_.1	0.572493	0.603834	0.451809	0.574472	0.561053	0.581758	0.510774	0.553997	0.519034	0.647074	...	0.593898	0.389431	0.254682	0.662202	0.435418	0.615406	0.234147	0.234147	0.254682	C5
TCGA.CS.4938.LGG.C5_.1	0.551088	0.617728	0.494956	0.608151	0.560231	0.548329	0.523150	0.536608	0.506502	0.629369	...	0.600972	0.415060	0.305500	0.659040	0.454508	0.611320	0.305500	0.223255	0.241685	C5
TCGA.E1.A7YU.LGG.C5_.1	0.531447	0.600096	0.479767	0.554191	0.562935	0.591940	0.550486	0.525981	0.492034	0.630475	...	0.590782	0.410158	0.280933	0.649194	0.469480	0.627374	0.295937	0.224542	0.271461	C5
TCGA.TM.A7C4.LGG.C5_.1	0.557283	0.631137	0.486160	0.559143	0.545856	0.556881	0.490815	0.503865	0.507647	0.646273	...	0.576507	0.429955	0.271722	0.669425	0.490772	0.626892	0.303506	0.222642	0.374109	C5
TCGA.HT.7483.LGG.C5_.1	0.550164	0.658776	0.556542	0.588742	0.559105	0.545667	0.490034	0.481184	0.470954	0.646835	...	0.600827	0.360948	0.175140	0.655678	0.483866	0.622553	0.289119	0.175140	0.348325	C5
TCGA.OR.A5J8.ACC.C6_.1	0.523671	0.575966	0.571506	0.523788	0.540236	0.512294	0.516848	0.466269	0.465364	0.594570	...	0.597697	0.253930	0.150130	0.591696	0.500395	0.670649	0.370831	0.150130	0.169537	C6
TCGA.D7.A748.STAD.C6_.1	0.506645	0.667325	0.510963	0.512853	0.469270	0.522373	0.603651	0.505210	0.420392	0.706066	...	0.594324	0.248820	0.097389	0.573912	0.459230	0.644050	0.290445	0.155997	0.127740	C6
TCGA.DX.AB2L.SARC.C6_.1	0.530221	0.603020	0.569074	0.559047	0.515473	0.571066	0.438389	0.449498	0.456640	0.491341	...	0.610022	0.460631	0.186813	0.601006	0.493623	0.663616	0.287781	0.356373	0.215465	C6
TCGA.BH.A0H5.BRCA.C6_.1	0.524347	0.554779	0.521978	0.552790	0.461483	0.501685	0.514258	0.487895	0.442896	0.586997	...	0.632985	0.191194	0.154078	0.562375	0.461398	0.635343	0.356803	0.066221	0.128751	C6
TCGA.D8.A141.BRCA.C6_.1	0.536416	0.601544	0.491975	0.578207	0.477198	0.478782	0.519373	0.485195	0.430024	0.555809	...	0.620485	0.345195	0.146276	0.576920	0.475919	0.636768	0.470924	0.090656	0.090656	C6
TCGA.A7.A0DB.BRCA.C6_.1	0.573799	0.611509	0.516464	0.507130	0.519182	0.559605	0.519199	0.449756	0.452602	0.640159	...	0.563253	0.313447	0.169471	0.591349	0.495301	0.635466	0.469915	0.150174	0.150174	C6
TCGA.FB.A4P5.PAAD.C6_.1	0.503231	0.601942	0.591569	0.532252	0.479578	0.509801	0.560365	0.465907	0.437401	0.551163	...	0.603698	0.288770	0.066713	0.560555	0.447937	0.666365	0.373320	0.066713	0.066713	C6
TCGA.FB.AAQ1.PAAD.C6_.1	0.547088	0.669354	0.556374	0.575363	0.511425	0.534338	0.584273	0.474198	0.450351	0.575780	...	0.548372	0.301072	0.202441	0.554637	0.472689	0.653401	0.307704	0.332028	0.125693	C6
TCGA.A8.A08Z.BRCA.C6_.1	0.564599	0.631282	0.543308	0.579381	0.499691	0.487058	0.551833	0.502932	0.450973	0.619468	...	0.592347	0.326527	0.180855	0.583227	0.506346	0.636528	0.490175	0.121022	0.121022	C6
TCGA.AR.A0TT.BRCA.C6_.1	0.550148	0.589376	0.496296	0.567811	0.494039	0.456793	0.514400	0.495892	0.408519	0.605587	...	0.614017	0.257182	0.137421	0.593452	0.497730	0.628467	0.509683	0.060906	0.137421	C6
TCGA.EL.A3GP.THCA.C6_.1	0.539347	0.602059	0.587730	0.534294	0.484431	0.492027	0.570387	0.489925	0.446972	0.655709	...	0.593318	0.306822	0.130648	0.598497	0.452201	0.624234	0.378650	0.130648	0.196287	C6
TCGA.DX.AB2Z.SARC.C6_.1	0.566796	0.606668	0.535313	0.563848	0.486424	0.468758	0.470592	0.483833	0.420579	0.597533	...	0.622022	0.303047	0.174420	0.565077	0.529220	0.654313	0.312090	0.135493	0.135493	C6
TCGA.05.5715.LUAD.C6_.1	0.526272	0.679035	0.496053	0.500599	0.496956	0.542688	0.579221	0.506409	0.479734	0.707386	...	0.549089	0.246295	0.198633	0.598890	0.479822	0.633190	0.211564	0.124307	0.124307	C6
TCGA.XF.AAN8.BLCA.C6_.1	0.490200	0.571726	0.679305	0.501556	0.500526	0.546044	0.551939	0.448028	0.427187	0.536573	...	0.579951	0.323108	0.166204	0.563020	0.430004	0.677198	0.331032	0.166204	0.166204	C6
TCGA.05.5715.LUAD.C6_.2	0.526272	0.679035	0.496053	0.500599	0.496956	0.542688	0.579221	0.506409	0.479734	0.707386	...	0.549089	0.246295	0.198633	0.598890	0.479822	0.633190	0.211564	0.124307	0.124307	C6
TCGA.3U.A98H.MESO.C6_.1	0.507051	0.607413	0.509807	0.548012	0.491548	0.550344	0.538366	0.490052	0.446313	0.493152	...	0.577190	0.336248	0.126303	0.551038	0.451374	0.676210	0.559915	0.098843	0.098843	C6
TCGA.E7.A7DU.BLCA.C6_.1	0.595040	0.595816	0.555534	0.475091	0.540981	0.602707	0.409030	0.500623	0.532144	0.507079	...	0.503172	0.354811	0.291779	0.617974	0.539230	0.669733	0.549867	0.241214	0.241214	C6
TCGA.E7.A7DU.BLCA.C6_.2	0.595040	0.595816	0.555534	0.475091	0.540981	0.602707	0.409030	0.500623	0.532144	0.507079	...	0.503172	0.354811	0.291779	0.617974	0.539230	0.669733	0.549867	0.241214	0.241214	C6
TCGA.A7.A0DB.BRCA.C6_.2	0.573799	0.611509	0.516464	0.507130	0.519182	0.559605	0.519199	0.449756	0.452602	0.640159	...	0.563253	0.313447	0.169471	0.591349	0.495301	0.635466	0.469915	0.150174	0.150174	C6
TCGA.FB.AAQ1.PAAD.C6_.2	0.547088	0.669354	0.556374	0.575363	0.511425	0.534338	0.584273	0.474198	0.450351	0.575780	...	0.548372	0.301072	0.202441	0.554637	0.472689	0.653401	0.307704	0.332028	0.125693	C6
TCGA.DX.A1KU.SARC.C6_.1	0.549936	0.652695	0.674654	0.521712	0.510103	0.501352	0.521481	0.501617	0.430390	0.517205	...	0.636496	0.431338	0.187093	0.586367	0.486879	0.666805	0.485532	0.163456	0.216325	C6
TCGA.3U.A98I.MESO.C6_.1	0.504295	0.620087	0.602403	0.542429	0.476758	0.503558	0.511589	0.457273	0.436591	0.539357	...	0.585780	0.476789	0.182592	0.563271	0.493496	0.659347	0.204457	0.146536	0.182592	C6

	ACTL6A_S5	ADAM9_S2	ADAMTS1_S5	ADCY7_S3	AIMP2_S5	ALKBH7_S5	ALOX5AP_S3	AMPD3_S3	APITD1_S5	APOC1_S3	...	WIPF1_S3	WNT2B_S2	WNT8B_S2	WSB2_S5	ZWILCH_S5	ZYX_S2	MMP3_S2	PLG_S5	RGS8_S5	Subgroup
TCGA.HF.7136.STAD.C1_	0.592349	0.628634	0.570785	0.536637	0.509823	0.462218	0.443087	0.482923	0.418519	0.483556	...	0.557671	0.348506	0.210016	0.629423	0.515653	0.664831	0.548405	0.144132	0.156693	C1
TCGA.EW.A1OZ.BRCA.C1_	0.616708	0.589806	0.546915	0.522103	0.523893	0.545416	0.518862	0.428246	0.447287	0.588167	...	0.543131	0.271636	0.215562	0.588503	0.520022	0.661331	0.488972	0.129630	0.147665	C1
TCGA.EY.A1GL.UCEC.C1_	0.591583	0.601375	0.530573	0.470164	0.573793	0.560062	0.408029	0.529847	0.516928	0.500291	...	0.537821	0.370267	0.256229	0.626342	0.500481	0.650543	0.602704	0.236607	0.236607	C1
TCGA.EB.A5FP.SKCM.C1_	0.604189	0.586198	0.566644	0.523076	0.575625	0.497918	0.362422	0.428469	0.544391	0.580780	...	0.635673	0.360680	0.261153	0.612561	0.564331	0.620690	0.457935	0.261153	0.261153	C1
TCGA.D1.A16D.UCEC.C1_	0.545903	0.632497	0.587917	0.533486	0.540958	0.556670	0.518468	0.510708	0.488154	0.585870	...	0.533510	0.332906	0.129106	0.609040	0.482094	0.670960	0.389138	0.282479	0.151172	C1
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
TCGA.43.2581.LUSC.C6_	0.544662	0.620647	0.575653	0.557788	0.515838	0.488698	0.592145	0.510097	0.408028	0.572360	...	0.584858	0.256580	0.154807	0.574657	0.465805	0.654632	0.372903	0.098287	0.098287	C6
TCGA.SH.A7BC.MESO.C6_	0.527214	0.635277	0.626745	0.563070	0.503132	0.523583	0.550000	0.501574	0.417098	0.594076	...	0.580386	0.560368	0.185317	0.567794	0.471550	0.683370	0.354513	0.093081	0.093081	C6
TCGA.EL.A3T9.THCA.C6_	0.538555	0.600622	0.599145	0.549116	0.533807	0.514959	0.532856	0.477059	0.474861	0.471344	...	0.587950	0.286881	0.189631	0.562508	0.489268	0.645087	0.247726	0.172708	0.172708	C6
TCGA.A2.A4S1.BRCA.C6_	0.513619	0.599382	0.589908	0.543182	0.478895	0.507401	0.570468	0.483158	0.441637	0.463836	...	0.601901	0.340524	0.171785	0.541964	0.450655	0.643585	0.259902	0.171785	0.171785	C6
TCGA.AR.A2LJ.BRCA.C6_	0.540358	0.551918	0.526153	0.489545	0.518348	0.603053	0.517177	0.453269	0.484431	0.650400	...	0.563845	0.215693	0.201434	0.558991	0.462621	0.676615	0.476988	0.133939	0.133939	C6

ICW1¶

The Project Questions¶

Data Preparation and Addressing Missingness¶

a. Data Cleaning¶

check the data type¶

remove the duplicates¶

finding outliers¶

b. Missing Data Imputation¶

check missing data¶

Imputation¶

Using IsolationForest Algortims to Estimate Outliers¶

c. Normalisation¶

Non-negative Matrix Factorization(NMF)¶

a. Application of NMF¶

b. Visualising¶

Principal Component Analysis(PCA)¶

a. Application of PCA¶

b. Plot and determining the number of retained components¶

3D PCA Visualisation¶

t-SNE Dimensionality Reduction¶

a. Application of t-SNE¶

b. Visualising t-SNE Results¶

¶

conclusion¶

	ACTL6A_S5	ADAM9_S2	ADAMTS1_S5	ADCY7_S3	AIMP2_S5	ALKBH7_S5	ALOX5AP_S3	AMPD3_S3	APITD1_S5	APOC1_S3	...	WIPF1_S3	WNT2B_S2	WNT8B_S2	WSB2_S5	ZWILCH_S5	ZYX_S2	MMP3_S2	PLG_S5	RGS8_S5	Subgroup
count	588	588	588	588	588	588	588	588	588	588	...	588	588	588	588	588	588	588	588	588	588
unique	540	540	540	540	540	540	540	540	540	540	...	540	540	540	540	540	540	540	540	540	6
top	0.547088474	0.669353783	0.556373656	0.575362692	0.511424709	0.534338128	0.584272643	0.474198022	0.450351433	0.575779881	...	0.548371597	0.301071951	0.202440846	0.554636793	0.47268884	0.653400943	0.307704021	0.332027894	0.125693204	C1
freq	3	3	3	3	3	3	3	3	3	3	...	3	3	3	3	3	3	3	3	3	98

	ACTL6A_S5	ADAM9_S2	ADAMTS1_S5	ADCY7_S3	AIMP2_S5	ALKBH7_S5	ALOX5AP_S3	AMPD3_S3	APITD1_S5	APOC1_S3	...	WDR77_S5	WIPF1_S3	WNT2B_S2	WNT8B_S2	WSB2_S5	ZWILCH_S5	ZYX_S2	MMP3_S2	PLG_S5	RGS8_S5
count	540.000000	540.000000	540.000000	540.000000	540.000000	540.000000	540.000000	540.000000	540.000000	540.000000	...	540.000000	540.000000	540.000000	540.000000	540.000000	540.000000	540.000000	540.000000	540.000000	540.000000
mean	0.562063	0.617497	0.549052	0.530075	0.532300	0.550593	0.482367	0.494148	0.478304	0.571891	...	0.550695	0.567491	0.356629	0.210863	0.607607	0.494693	0.643181	0.320501	0.220009	0.191605
std	0.031346	0.046260	0.058148	0.039704	0.037663	0.045730	0.053319	0.044860	0.036284	0.087900	...	0.030780	0.036571	0.074973	0.063109	0.034481	0.035625	0.026520	0.130704	0.148883	0.078200
min	0.454512	0.481747	0.367899	0.420165	0.394824	0.406076	0.331671	0.368469	0.322530	0.324714	...	0.469474	0.445756	0.175897	0.055078	0.491753	0.393490	0.542554	0.056857	0.026909	0.046179
25%	0.541103	0.594717	0.509507	0.501417	0.505232	0.519779	0.444551	0.465861	0.452666	0.514803	...	0.527979	0.541440	0.304264	0.164997	0.581120	0.468695	0.626757	0.220523	0.138133	0.134756
50%	0.559485	0.618027	0.544527	0.531558	0.535666	0.552652	0.484568	0.492815	0.478269	0.570587	...	0.551934	0.571641	0.351196	0.209848	0.603842	0.493034	0.641369	0.286077	0.186938	0.178872
75%	0.581467	0.644837	0.586762	0.559147	0.560231	0.582331	0.520911	0.517119	0.501981	0.619868	...	0.572067	0.595177	0.399560	0.253092	0.633591	0.519085	0.660907	0.418848	0.232155	0.237335
max	0.710251	0.746323	0.729066	0.678374	0.636284	0.679299	0.616650	0.682606	0.569738	0.908518	...	0.648555	0.666154	0.611041	0.455048	0.704738	0.613592	0.745479	0.782070	0.833178	0.540641

	C1orf54_S3	HN1L_S5	PTPRC_S3	SAMHD1_S3
TCGA.HF.7136.STAD.C1_	0.364874	0.611415	0.532904	0.530763
TCGA.EW.A1OZ.BRCA.C1_	0.421521	0.641031	0.525117	0.533921
TCGA.EY.A1GL.UCEC.C1_	0.457814	0.637443	0.424625	0.486925
TCGA.EB.A5FP.SKCM.C1_	0.427228	0.642142	NaN	0.463744
TCGA.D1.A16D.UCEC.C1_	0.398250	0.700416	0.493172	0.554316
...	...	...	...	...
TCGA.43.2581.LUSC.C6_	0.351794	0.599456	NaN	0.591878
TCGA.SH.A7BC.MESO.C6_	0.425960	0.614321	0.582419	0.588247
TCGA.EL.A3T9.THCA.C6_	0.449813	0.602357	0.531815	0.544002
TCGA.A2.A4S1.BRCA.C6_	NaN	0.555404	0.533358	0.511731
TCGA.AR.A2LJ.BRCA.C6_	0.465609	0.616445	0.529782	0.515932

	ACTL6A_S5	ADAM9_S2	ADAMTS1_S5	ADCY7_S3	AIMP2_S5	ALKBH7_S5	ALOX5AP_S3	AMPD3_S3	APITD1_S5	APOC1_S3	...	WIPF1_S3	WNT2B_S2	WNT8B_S2	WSB2_S5	ZWILCH_S5	ZYX_S2	MMP3_S2	PLG_S5	RGS8_S5	Subgroup
TCGA.HF.7136.STAD.C1_	0.464206	0.553018	0.550431	0.422275	0.401367	0.176712	0.390964	0.364345	0.395560	0.295254	...	0.553260	0.396224	0.387375	0.646382	0.555027	0.510994	0.677798	0.145389	0.223503	C1
TCGA.EW.A1OZ.BRCA.C1_	0.574901	0.405547	0.482630	0.363034	0.470152	0.514156	0.656861	0.190290	0.514111	0.489703	...	0.481381	0.219441	0.401239	0.454258	0.574876	0.489773	0.595846	0.127402	0.205245	C1
TCGA.EY.A1GL.UCEC.C1_	0.460726	0.449487	0.436215	0.151328	0.714113	0.573561	0.267944	0.513717	0.801093	0.326360	...	0.455133	0.446272	0.502915	0.631917	0.486097	0.424361	0.752671	0.260084	0.385122	C1
TCGA.D1.A16D.UCEC.C1_	0.253135	0.567689	0.599091	0.409430	0.553581	0.559803	0.655478	0.452793	0.682517	0.485432	...	0.433820	0.360347	0.185082	0.550679	0.402559	0.548159	0.458184	0.316978	0.212338	C1
TCGA.R6.A8W8.ESCA.C1_	0.518248	0.760312	0.402481	0.550812	0.499165	0.363376	0.327758	0.303664	0.745557	0.228929	...	0.387164	0.326819	0.479786	0.498336	0.794955	0.126130	0.707623	0.197941	0.278473	C1
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
TCGA.43.2581.LUSC.C6_	0.247497	0.522681	0.564256	0.508487	0.430773	0.284113	0.914014	0.450848	0.352329	0.460321	...	0.687661	0.184814	0.249341	0.389247	0.328550	0.449154	0.435797	0.088529	0.105384	C6
TCGA.SH.A7BC.MESO.C6_	0.168205	0.578246	0.709375	0.530017	0.368653	0.425604	0.766126	0.423715	0.389704	0.500686	...	0.665557	0.883463	0.325622	0.357024	0.354655	0.623405	0.410439	0.082071	0.094855	C6
TCGA.EL.A3T9.THCA.C6_	0.219743	0.446626	0.630981	0.473138	0.518621	0.390625	0.705965	0.345678	0.627741	0.272554	...	0.702948	0.254499	0.336407	0.332205	0.435153	0.391279	0.263191	0.180832	0.255893	C6
TCGA.A2.A4S1.BRCA.C6_	0.106425	0.441916	0.604747	0.448950	0.250162	0.359972	0.837949	0.365091	0.490826	0.258597	...	0.771916	0.377868	0.291789	0.235748	0.259721	0.382173	0.279981	0.179686	0.254025	C6
TCGA.AR.A2LJ.BRCA.C6_	0.227935	0.261649	0.423659	0.230325	0.443044	0.747928	0.650948	0.269944	0.667177	0.605381	...	0.583781	0.090782	0.365918	0.315691	0.314086	0.582450	0.579321	0.132746	0.177485	C6

	TCGA.HF.7136.STAD.C1_	TCGA.EW.A1OZ.BRCA.C1_	TCGA.EY.A1GL.UCEC.C1_	TCGA.D1.A16D.UCEC.C1_	TCGA.R6.A8W8.ESCA.C1_	TCGA.CM.6676.COAD.C1_	TCGA.L5.A43C.ESCA.C1_	TCGA.31.1944.OV.C1_	TCGA.D5.6920.COAD.C1_	TCGA.K4.A3WS.BLCA.C1_	...	TCGA.55.7281.LUAD.C6_	TCGA.85.8049.LUSC.C6_	TCGA.FX.A2QS.SARC.C6_	TCGA.CG.4462.STAD.C6_	TCGA.IB.7889.PAAD.C6_	TCGA.43.2581.LUSC.C6_	TCGA.SH.A7BC.MESO.C6_	TCGA.EL.A3T9.THCA.C6_	TCGA.A2.A4S1.BRCA.C6_	TCGA.AR.A2LJ.BRCA.C6_
0	0.670273	0.536199	0.738639	0.570018	0.773193	0.825512	0.253430	0.655150	0.700875	0.518160	...	0.213521	0.267566	0.366987	0.079072	0.272852	0.205031	0.140498	0.403977	0.182884	0.314995
1	0.374405	0.436802	0.249843	0.415064	0.345064	0.205967	0.707069	0.354750	0.392410	0.531307	...	0.697346	0.739991	0.457509	0.768106	0.660736	0.700451	0.660060	0.513931	0.615291	0.587001
2	0.043526	0.130171	0.122499	0.136425	0.000000	0.101441	0.074782	0.107439	0.005845	0.000000	...	0.160447	0.000583	0.285564	0.200387	0.123943	0.160314	0.281538	0.172972	0.290966	0.187873

	TCGA.HF.7136.STAD.C1_	TCGA.EW.A1OZ.BRCA.C1_	TCGA.EY.A1GL.UCEC.C1_	TCGA.D1.A16D.UCEC.C1_	TCGA.R6.A8W8.ESCA.C1_	TCGA.CM.6676.COAD.C1_	TCGA.L5.A43C.ESCA.C1_	TCGA.31.1944.OV.C1_	TCGA.D5.6920.COAD.C1_	TCGA.K4.A3WS.BLCA.C1_	...	TCGA.55.7281.LUAD.C6_	TCGA.85.8049.LUSC.C6_	TCGA.FX.A2QS.SARC.C6_	TCGA.CG.4462.STAD.C6_	TCGA.IB.7889.PAAD.C6_	TCGA.43.2581.LUSC.C6_	TCGA.SH.A7BC.MESO.C6_	TCGA.EL.A3T9.THCA.C6_	TCGA.A2.A4S1.BRCA.C6_	TCGA.AR.A2LJ.BRCA.C6_
0	0.214136	0.422697	0.305488	0.277269	0.211139	0.209205	0.000000	0.275706	0.110916	0.081141	...	0.009077	0.000000	0.082434	0.000000	0.044881	0.053565	0.000000	0.169299	0.000000	0.138789
1	0.140067	0.287877	0.016342	0.260257	0.159873	0.015294	0.488001	0.196189	0.213669	0.194430	...	0.546109	0.481025	0.241396	0.584710	0.472232	0.599988	0.496395	0.281465	0.410616	0.453305
2	0.024846	0.000000	0.043107	0.012522	0.000000	0.069041	0.000000	0.000000	0.000000	0.000000	...	0.013214	0.000000	0.183095	0.029964	0.000000	0.000000	0.131154	0.050251	0.081151	0.000000
3	0.000000	0.334069	0.406192	0.355979	0.000000	0.260131	0.050107	0.530276	0.104389	0.354174	...	0.347402	0.065720	0.076648	0.000000	0.281784	0.102888	0.223301	0.282730	0.402765	0.590644
4	0.514916	0.108418	0.190731	0.147809	0.692668	0.563621	0.266254	0.186807	0.654390	0.175588	...	0.038262	0.218572	0.289766	0.099579	0.072575	0.236687	0.036853	0.021897	0.000000	0.000000
5	0.378831	0.415083	0.213826	0.070757	0.071210	0.182870	0.366757	0.213738	0.089264	0.314985	...	0.164927	0.302884	0.609481	0.413940	0.073662	0.360834	0.345360	0.347544	0.418344	0.176426
6	0.015444	0.189833	0.000000	0.074263	0.183608	0.000000	0.026205	0.090180	0.042874	0.178225	...	0.107719	0.099827	0.200752	0.019641	0.099120	0.105352	0.157319	0.026095	0.147863	0.109185
7	0.326869	0.000000	0.410831	0.470911	0.329749	0.255936	0.222796	0.270383	0.292043	0.137576	...	0.430531	0.272135	0.000000	0.271405	0.320940	0.287293	0.272190	0.428403	0.106149	0.008776
8	0.089061	0.000000	0.200452	0.120332	0.141949	0.239100	0.156547	0.055248	0.107852	0.065014	...	0.008742	0.000000	0.227771	0.301777	0.218083	0.012661	0.086827	0.201864	0.386798	0.087923
9	0.293239	0.221573	0.334435	0.365959	0.356479	0.299661	0.465909	0.239568	0.488739	0.601135	...	0.388456	0.561803	0.000000	0.359692	0.608885	0.176441	0.203119	0.282571	0.158013	0.541271

	ACTL6A_S5	ADAM9_S2	ADAMTS1_S5	ADCY7_S3	AIMP2_S5	ALKBH7_S5	ALOX5AP_S3	AMPD3_S3	APITD1_S5	APOC1_S3	...	WDR77_S5	WIPF1_S3	WNT2B_S2	WNT8B_S2	WSB2_S5	ZWILCH_S5	ZYX_S2	MMP3_S2	PLG_S5	RGS8_S5
0	0.523795	0.386598	0.413939	0.179554	0.683130	0.656076	0.123648	0.318118	0.748332	0.260790	...	0.687595	0.160825	0.361301	0.537607	0.533319	0.680243	0.340690	0.430017	0.331779	0.321591
1	0.118208	0.598067	0.594447	0.465999	0.077371	0.191279	0.856451	0.351218	0.380590	0.506687	...	0.062096	0.870081	0.341793	0.032127	0.252569	0.221784	0.447574	0.394233	0.074312	0.000000
2	0.163163	0.392037	0.269655	0.451391	0.554604	0.525198	0.489633	0.400525	0.512428	0.466239	...	0.384655	0.628547	0.392321	0.424888	0.665708	0.229681	0.207330	0.089724	0.194502	0.464619

	ACTL6A_S5	ADAM9_S2	ADAMTS1_S5	ADCY7_S3	AIMP2_S5	ALKBH7_S5	ALOX5AP_S3	AMPD3_S3	APITD1_S5	APOC1_S3	...	WDR77_S5	WIPF1_S3	WNT2B_S2	WNT8B_S2	WSB2_S5	ZWILCH_S5	ZYX_S2	MMP3_S2	PLG_S5	RGS8_S5
0	0.430239	0.379590	0.146470	0.000000	0.529449	0.431899	0.101489	0.326058	0.577687	0.252097	...	0.471757	0.045296	0.028522	0.516410	0.540310	0.490225	0.000000	0.116208	0.186179	0.317912
1	0.038995	0.337747	0.000000	0.291349	0.071195	0.037884	0.840680	0.201924	0.145153	0.647377	...	0.028383	0.598717	0.000000	0.000000	0.156235	0.054440	0.077668	0.000000	0.000000	0.000000
2	0.132580	0.401605	0.291162	0.544452	0.532018	0.490208	0.462853	0.422952	0.462383	0.380538	...	0.359222	0.693346	0.483211	0.416484	0.695047	0.192911	0.217656	0.134679	0.151318	0.498077
3	0.148231	0.000000	0.201518	0.000000	0.397999	0.620760	0.214920	0.081995	0.480424	0.311009	...	0.456640	0.037858	0.214072	0.221557	0.169033	0.258734	0.441062	0.078630	0.099786	0.121079
4	0.523253	0.252605	0.000000	0.286569	0.514471	0.175233	0.007627	0.042139	0.267296	0.113498	...	0.549039	0.080647	0.285511	0.273995	0.263097	0.706149	0.153551	0.467293	0.144149	0.100824
5	0.163038	0.373766	0.575184	0.335221	0.061498	0.000000	0.287167	0.031794	0.121129	0.000000	...	0.000000	0.544400	0.439029	0.087329	0.105917	0.424607	0.412070	0.587580	0.000000	0.105456
6	0.132902	0.000000	0.074577	0.199110	0.000000	0.000000	0.079860	0.097504	0.193008	0.114607	...	0.000000	0.161817	0.015076	0.000000	0.101570	0.150573	0.000000	0.124582	0.197832	0.000000
7	0.132385	0.654977	0.415638	0.382127	0.159397	0.131819	0.480442	0.551267	0.459787	0.000000	...	0.224082	0.144927	0.263215	0.096001	0.328916	0.056305	0.227068	0.179336	0.000000	0.154254
8	0.000000	0.000000	0.298541	0.000000	0.220628	0.306916	0.000000	0.009732	0.251184	0.268957	...	0.051885	0.210917	0.076694	0.139759	0.163678	0.019450	0.128711	0.000000	0.435797	0.152734
9	0.000000	0.183018	0.403738	0.077339	0.000000	0.258730	0.041755	0.188054	0.136978	0.000000	...	0.056312	0.391260	0.290523	0.118597	0.151052	0.007205	0.205977	0.253172	0.000000	0.000000

	PC1	PC2	PC3
TCGA.HF.7136.STAD.C1_	-0.220870	-1.573532	1.627781
TCGA.EW.A1OZ.BRCA.C1_	-0.650727	-0.771769	0.467953
TCGA.EY.A1GL.UCEC.C1_	0.828764	-1.788312	-0.247809
TCGA.D1.A16D.UCEC.C1_	-0.403537	-0.892226	-0.592516
TCGA.R6.A8W8.ESCA.C1_	0.086907	-2.157220	1.319764
...	...	...	...
TCGA.43.2581.LUSC.C6_	-2.844417	0.868858	0.636143
TCGA.SH.A7BC.MESO.C6_	-2.477495	1.364205	0.053401
TCGA.EL.A3T9.THCA.C6_	-1.301796	-0.075522	-0.139321
TCGA.A2.A4S1.BRCA.C6_	-2.109095	1.165129	-0.796590
TCGA.AR.A2LJ.BRCA.C6_	-1.866846	0.395933	-1.379289

	feature1	feature2	feature3
TCGA.HF.7136.STAD.C1_	-1.614262	-11.426699	-5.145396
TCGA.EW.A1OZ.BRCA.C1_	-1.041307	-1.546486	-3.651807
TCGA.EY.A1GL.UCEC.C1_	4.791106	-4.639829	-1.500908
TCGA.D1.A16D.UCEC.C1_	0.634710	-5.939110	-0.696586
TCGA.R6.A8W8.ESCA.C1_	0.456902	-12.938315	-6.142682
...	...	...	...
TCGA.43.2581.LUSC.C6_	-13.882877	-0.772022	3.460467
TCGA.SH.A7BC.MESO.C6_	-9.458762	-0.729669	-4.345543
TCGA.EL.A3T9.THCA.C6_	-5.380478	0.647410	3.277821
TCGA.A2.A4S1.BRCA.C6_	-8.747445	2.095522	-5.790107
TCGA.AR.A2LJ.BRCA.C6_	-4.704082	-1.634271	-2.809523

	feature1	feature2
TCGA.HF.7136.STAD.C1_	9.009348	-15.098787
TCGA.EW.A1OZ.BRCA.C1_	-7.607230	-14.919571
TCGA.EY.A1GL.UCEC.C1_	9.072879	-2.854402
TCGA.D1.A16D.UCEC.C1_	-4.740012	-4.195903
TCGA.R6.A8W8.ESCA.C1_	9.611462	-13.326759
...	...	...
TCGA.43.2581.LUSC.C6_	-18.469227	-14.071336
TCGA.SH.A7BC.MESO.C6_	-22.456432	-6.157858
TCGA.EL.A3T9.THCA.C6_	-15.806913	-4.879653
TCGA.A2.A4S1.BRCA.C6_	-21.192274	-1.086885
TCGA.AR.A2LJ.BRCA.C6_	-19.546152	-6.487153

	feature1	feature2	Subgroup
TCGA.HF.7136.STAD.C1_	9.009348	-15.098787	C1
TCGA.EW.A1OZ.BRCA.C1_	-7.607230	-14.919571	C1
TCGA.EY.A1GL.UCEC.C1_	9.072879	-2.854402	C1
TCGA.D1.A16D.UCEC.C1_	-4.740012	-4.195903	C1
TCGA.R6.A8W8.ESCA.C1_	9.611462	-13.326759	C1
...	...	...	...
TCGA.43.2581.LUSC.C6_	-18.469227	-14.071336	C6
TCGA.SH.A7BC.MESO.C6_	-22.456432	-6.157858	C6
TCGA.EL.A3T9.THCA.C6_	-15.806913	-4.879653	C6
TCGA.A2.A4S1.BRCA.C6_	-21.192274	-1.086885	C6
TCGA.AR.A2LJ.BRCA.C6_	-19.546152	-6.487153	C6